Review: Parallax-Audio – VirtualSoundStage
We take a close look at Virtual Sound Stage, a tool for placing sounds in two dimensions (left-to-right and front-to-back) in your mix’s audio space.
by Per Lichtman, May 2013
[Update: Parallax Audio has released Virtual Sound Stage 2 and we’ve gotten a review copy. Keep an eye out for the review in the next issue, coming in January, 2014.]
VirtualSoundStage is a sound placement tool, used to visually pan and place different sounds, voices or instruments across the soundstage (both left to right and front to back) with an emphasis on use in music and an interface designed to be used by musicians as opposed to engineers. The combination of approach, CPU usage and price separates it from other products so keep reading to find out the details.
As of May 2013, VirtualSoundStage costs $99 USD and available as download only exclusively from http://www.parallax-audio.com/. This means it is much less expensive than every competing plug-in I’ve encountered so far. The cheapest other options I’ve found are at least 50% more (with most others costing much more) and no competitor close to it in price offers a similarly simple interface with predictable CPU usage.
Ease of Use and Functionality
VirtualSoundStage (VSS) is very easy to use. It takes longer to describe than to just click and use it: you just click on a speaker icon and drag to where you want to hear the sound coming from on a diagram. If you are panning sounds that are originally centered or mono and like the default settings and, that might be all you need, but I’ll talk about it a bit more in-depth just in case. Of course, there’s so much great information in the well-written manual that you’d have to be crazy to pass on reading that, too – especially the appendices. 🙂
The bulk of the work is done with a large diagram that shows the walls of the virtual soundstage, a suggested seating chart (generally speaking), a conductor graphic (to give a sense of where your ears are on the stage) and most importantly a speaker that shows where your sound is coming from. You just click and drag the speaker icon to put the sound wherever you like (or if you would rather use knobs, you can do the same thing using the “angle” and “distance” knobs below it).
One of the reasons that VSS is just as good for advanced users as for beginners is that it gives you lots of options for inputting changes without getting cluttered or overwhelming. Want to change how far away an instrument sounds? You can drag it on the diagram, change the value on a knob or double-click on the knob to bring up a dialog box to type in the value (from a minimum of 1 to up to 13 to 20 meters, depending on angle you are using)
VSS does not specifically address width the way that some competing products do – you either pan the full stereo spread or you use Monomaker in the Input Offset menu to choose the left or right channel. While the monomaker is helpful and saves you from having to use a separate plug-in to choose just one channel from a stereo signal to narrow the stereo field, you may want to look into supplementing this functionality with other products if it is important to you (though that doesn’t have to mean spending a lot of money as basic functionality can be added with free products like dfx Monomaker).
Whether you opt to emphasize VirtualSoundStage’s early reflections or choose to bring the slider all the way down, the results of using VirtualSoundStage sound different from using panning alone (and I found that was true whether I used the built-in panning in DAWs like Reaper and Reason as well as with dedicated panning plug-ins like Vienna Audio Suite PowerPan or Waves S-1). Nine times out of ten, I prefer the VSS result and there is a lot of room to configure it between the microphone blend (left, center and right faders), the early reflections to direct signal (two faders) and the air absorption filter. I found the sound held up well, even when I used over 190 instances in a single session (though that of course meant having to bounce on most systems).
What are some alternatives and how does it compare?
VirtualSoundStage is somewhat unusual in terms of its combination of approach and functionality. A handful of orchestral libraries include “stage position” style interfaces, but they are typically proprietary and limited to use with that library only. Visually placing sounds from any source you choose was previously only possible with more complicated and expensive products like IRCAM SPAT, IOSonos AnyMix or VSL MIR (all of which differ greatly in their respective functionality and focus). VSS is less expensive and a lot less intimidating to use. Compared to it’s most closely priced competition, WaveArts Panorama, the placement window is much larger, the interface less cluttered and easier to understand and more targeted towards musicians and composers than engineers – making it easier to use but also less flexible for sound design work or advanced engineering.
Unlike most normal panning tools, the sound does not update continuously as you adjust the distance and angle of your sound from the listener. There is a slight delay from when you let go of the settings until they get applied. The angle and distance also cannot be automated at the time of writing. This means that VSS is better for static positioning than any sounds that need to move across the stereo field dynamically.
In other words, while some of the other products can do things that VSS can’t, none of them make it easy to use or inexpensive for it’s target audience and VSS fills a very real gap in the market.
I tested VSS on 3 computers. Two older laptops with 4GB of RAM (one a Windows Vista 64-bit PC and the other a Mountain Lion MacBook Pro) as well as a newer desktop running Windows 7 64-bit with 8GB of RAM and a 6-core AMD 1090T processor. Performance was consistent, stable and light on CPU.
For example, under Mountain Lion with the 2.4 GHz Core 2 Duo in the 2010-era MacBook Pro I tested it at multiple samplerates in Cockos Reaper v4.402 (with anticipative FX disabled) I compared CPU usage with a 512 sample buffer with the internal sound and Core Audio.
At 96 KHz, each plug-in instance used less than 1% CPU based on the Reason Performance meter, even with the Air Absorption filter engaged. At 44.1 KHz each instance used less than 0.5% (in other words it would add less than 50% CPU usage if you added 100 instances). On the newer 6-core system CPU usage was even better and testing it seemed like a purely academic exercise.
For reference, I did some close comparisons with WaveArts Panorama and I found a big CPU usage difference. At 96 KHz, VSS used between 0.9% and 1.0% (mainly depending on whether the air absorption filter was engaged) – in fact the difference was small it was difficult to measure fully accurately. By comparison, WaveArts Panorama used between 0.5% (when in headphones mode, using a mono stereo field and only computing the direct signal) to 23.6% while automating parameters and using the direct and ER settings with the most CPU intensive settings, or 24.6% if the internal reverb tail was used (which to be fair is a feature that VSS does not model, though the quality here was not to my taste). I also did not find any air absorption filter options in WaveArts Panorama, so I could not compare that
In other words, CPU usage in VSS is both light and predictable (less than ⅕ variation between minimum and maximum CPU usage), while the additional sound design functionality of WaveArts Panorama means that CPU usage is almost 50x higher for the maximum settings from the minimum ones and the maximum usage is almost 25 times higher than for VSS.
No One’s Perfect
For every advantage VSS has over many competing produts (low CPU usage, low-price, easy to use, uncluttered interface) it also does not currently support automation for most parameters (including angle and distance), which some of those products do. This means that for advanced sound design (as opposed to simple orchestral placement) you might be better served by products more targeted towards this, such as WaveArts Panorama, IOSONOS Anymix or IRCAM SPAT
And if you want a more advanced reverb implementation based on a larger IR collection, then VSL’S MIR or MIR24 might be a better option. Be aware that all of the other products mentioned involve an increase in price and the few I’ve used also involve an increase in complexity and CPU usage.
As far as the other aspects of VSS, the least intuitive part of dealing with VirtualSoundStage are the Input Offset and Preset sections. Both of these require reading the manual to understand how to use them properly. As a novice user, opening up the advanced screen for Input Offset is a little bewildering: when do I use it and why? I also wish that the Input Offset screen were easier to use – at the moment you have to rely on external meters (like the free Voxengo SPAN) to check the balance while you adjust it. It would be great if metering (especially input and output metering side-by-side) were added in a future update as several competing products have at least some metering already.
In the case of the presets, the complication has less to do with the software design and more to do with the many possibilities. How do you know if the presets are designed to make other sounds match the positions used in the libraries mentioned vs. designed to place the libraries mentioned in locations that make sense? I had to read the manual to find out. But in all fairness, the manual is actually a remarkably good read and you should do yourself a favor and check out the appendices.
For the future
Their site lists plans for future updates, many of which would greatly speed up the process of working with VSS and make it even easier to use – even though it is already one of the simplest of it’s kind. You can read about these plans in the “What’s to Come” section of the manual currently posted at http://www.parallax-audio.com/index.php/documentation at the time of writing
My favorite things about VSS are how easy it is to use and how good it sounds – it’s difficult to go back to standard panning after you get spoiled by VSS’s approach. Of course the low CPU usage means that most of the time you don’t have to. Most of the things I wanted to when placing sounds were easy to figure out and accomplish quickly and easily. It is by far the most musician/composer targeted product of it’s kind the price-point (or up to three times the price-point).
So if positioning mono or stereo sounds in your mix is important to you, then I would highly recommend VSS – it’s a great addition to your tool set offering something different from every other placement tool in terms of it’s emphasis and approach and it sounds great. I hope to see them implement some of the features they’ve planned for the future, such as adding control for multiple instance in a single layout or a dedicated input offset plug-in. It would also be great to see metering or better automation support, but perhaps the simplest addition would be to add the information from “Seating Layout” appendix onto the seating chart diagram. But VSS is also already a great product in its current form, even without any updates.
If you leave everything panned center (for instance if you primarily record live events with a pair of stereo mics) then you don’t really benefit from a panning tool at all – so you might want to pass on buying one, even if it’s as good as VSS. 😉 But for musicians and composers that want a natural and intuitive way of placing their sounds that doesn’t require them to get deep into thinking like an engineer (or go deep into their wallet or CPU usage) VSS is quite simply the best tool available today.